ABSTRACT Major depressive disorder (MDD) is the world?s leading cause of disability. While effective treatments for MDD are available, ~50% of patients will not experience clinical benefits after 3-months of treatment. Of these patients, -30% are defined to have treatment-resistant depression (TRD). Acting to inhibit NMDA receptors, ketamine is shown to produce a rapid therapeutic response in up to 70% of TRD patients, though the antidepressant mechanisms themselves are not understood. While many studies have used neuroimaging to show that MDD is associated with particular brain abnormalities, few imaging studies have addressed the neural effects of ketamine and/or have determined whether imaging measures prior to treatment might dissociate individuals who will benefit from ketamine. By leveraging training opportunities and infrastructure provided by a larger funded project for which the sponsor of this application serves as PI, to address this gap, this F32 proposal uses diffusion MRI (dMRI) and resting-state functional MRI (rsfMRI) to determine whether serial ketamine therapy relates to changes in structural and functional connectivity within and between brain regions widely implicated in MDD. Patients (n=60) are scanned and receive mood ratings before treatment, 24 hours after the first ketamine infusion, and 24 hours and 6 weeks after last ketamine treatment. Demographically similar controls (n=40) receive MRI scans at two time points to establish normative values. Our first aim addresses whether white matter integrity measured with dMRI and functional connectivity measured with rsfMRI in fronto-limbic-striatal circuits at baseline relate to improved depressive symptoms 24 h after first and last ketamine infusion. We hypothesize that ketamine non-responders will have greater deficits in structural and functional connectivity in these circuits compared to responders and controls. Our second aim addresses whether ketamine therapy is associated with longitudinal changes in structural and functional connectivity in these same networks. We hypothesize that ketamine responders will show a normalization or compensation in network connectivity, with no change over time in non-responders and that connectivity measures will return to baseline values 6 weeks after treatment in patients who relapse. Since ketamine may benefit symptoms of anhedonia and apathy more than other treatments, Aim 3 will investigate the neural correlates of these behavioral constructs following serial ketamine and determine if changes are sustained at 6-weeks. We hypothesize ketamine will relate to pronounced reductions in anhedonia and apathy in responders and that degree of change will relate to changes in structural and functional connectivity; we will test whether these behavioral changes are more or less sustainable at 6-weeks. Research outcomes are expected to clarify the mechanisms underlying antidepressant response to ketamine therapy, which could impact future treatment decisions to benefit individual patients. This project will provide important training and career development opportunities in computational image analysis and MDD treatment research.